Im having problem to execute AUC, can someone help me, i read somewhere that perfcurve function could not be used for multi-class classification. I tried trapz but it is still error.
net = patternnet(hiddenLayerSize, 'trainscg');
net.divideParam.trainRatio = 70/100;
net.divideParam.valRatio = 15/100;
net.divideParam.testRatio = 15/100;
[net,tr] = train(net,inputs,targets);
e = gsubtract(targets,outputs);
performance = perform(net,targets,outputs)
[c,cm]=confusion(targets,outputs);
fprintf('Percentage Correct Classification: %f%%\n', (1-c)* 100);
fprintf('Percentage Incorrect Classification: %f%%\n', 100*c);
fprintf('Error performance: %f\n',performance)
figure, plotconfusion(targets,outputs)
figure, plotroc(targets,outputs)
[tpr,fpr] = roc(targets,outputs);
colAUC = perfcurve(fpr,tpr);
these are my code, ignore the %% Obtain the AUC section because it is not working.
Not enough input arguments.
Error in perfcurve (line 450)
posClass,negClass,trueNames);
colAUC = perfcurve(fpr,tpr);
and these are the error i got when run the command above.
